Why is a CNN translation invariant?
Translational Invariance makes the CNN invariant to translation. Invariance to translation means that if we translate the inputs the CNN will still be able to detect the class to which the input belongs. Translational Invariance is a result of the pooling operation.
What are the advantages of convolutional neural networks?
The main advantage of CNN compared to its predecessors is that it automatically detects the important features without any human supervision. For example, given many pictures of cats and dogs it learns distinctive features for each class by itself. CNN is also computationally efficient.
How are convolutional neural networks able to learn invariant features?
The convolution operation on images is translation equivariant. Sliding the image over will have the same output but simply translated over. Therefore, convolutional neural networks which are composed of convolutions, are also translation-equivariant. How is a convolutional neural network able to learn invariant features?
What makes a CNN invariant to object translation?
In conclusion, what makes a CNN invariant to object translation is the presence of convolution filters. Additionally, I believe that if a CNN is trained showing faces only at one corner during the learning process, the fully-connected layer may become insensitive to faces in other corners.
Do you need to change convolutional part of model?
It is interesting to note, that this does not require a change of the convolutional part of the model, as convolutions are independent of the input size of the image. It would, however, require a change of the fully connected output layer, which was not used for the visualization above.
What do you need to know about translation invariance?
Translation invariance requires, that the output of a mapping / network does not change when the input is translated. This is possible to achieve approximatively, up to a certain amount, but not directly. We will first investigate a related property, which is translation equivariance. What is an equivariant mapping?